This paper describes the proposal of "Frequency Range Method" by which metal surface temperature fluctuation can be evaluated using the effective heat transfer coefficient predictid by "the power spectrum method" and fluid temperature fluctuation data. The validity of the Frequency Range Method was examined using parallel impinging jet test data on the basis of comparison of "the Improved Time Range Method" developed by the author. It was found that the metal temperature fluctuations predicted by the two methods agreeded with the corresponding experimental data and the validity of the methods were demonstrated. Metal surface temperature fluctuation was predicted by the two methods. The predicted results appeared to be in good agreement each other and no marked difference in prediction accuracy were found between the Improved Time Range Method and the Frequency Range Method.